Aims/hypothesis. To analyse whether the time of diagnosis of coeliac disease with respect to the clinical onset of diabetes could differentiate subgroups of varying severity in patients with both diseases.\ud
\ud
Methods. We investigated 383 patients with Type I (insulin-dependent) diabetes mellitus for coeliac disease. Sex distribution, age at diagnosis of diabetes, prevalence of ketoacidosis at the onset of diabetes and prevalence of other autoimmune diseases were compared in patients. We divided these patients according to whether coeliac disease was diagnosed before (Group A, n=8) or after (Group B, n=24) diabetes onset and whether they had presented clinical symptoms of coeliac disease. Group C (n=351) included diabetic patients without coeliac disease.\ud
\ud
Results. Out of 383 Type I diabetic patients we found 32 coeliac subjects (8.3%). There was a higher number of girls (p=0.003), but similar age and prevalence of ketoacidosis compared with Group C; 18.7% had a third autoimmune disorder. The higher number of girls was confirmed in Groups A and B in comparison to Group C (p=0.013), while higher prevalence of both ketoacidosis (p=0.009) and other autoimmune diseases (p=0.001) was found only in Group A. Compared with symptomatic patients, asymptomatic subjects in Group B had a lower number of girls, older age at diabetes onset, lower prevalence of ketoacidosis and no other associated autoimmune disease.\ud
\ud
Conclusions/interpretation. A wide clinical spectrum characterises the association of coeliac disease and diabetes mellitus, with a severe clinical presentation (higher prevalence of ketoacidosis at the onset and occurrence of other autoimmune diseases) when coeliac disease is diagnosed before diabetes. Distinct phenotypes might imply the contribution of a peculiar genetic background